The modular robot

by @ulaulaman about #MIT #robot #technology #modular

Starting from a famous speech by Richard Feynman (pdf), it was born nanotechnology. One day in future we'll probably have nano-chip, nano-computer, nano-bot, but for now we must content ourselves with the robots with wheels and legs that explore the remote corners of the Earth, and also arriving on Mars in the past years.
A possible evolution of robots is described in science fiction: for example in Alan Moore's Tom Strong, one of the enemy of the hero is the Modular Man, an electronic entity consistuted by many memory modules, separated but which together realized one of the most advanced and deadliest artificial intelligence in the world. Using the Modular Man, Moore brings back a classic of the genre: the revolt of the technology against mankind. A variation of this idea is proposed by Frank Schatzing in The Swarm, but in this last case the superintelligent entity is constituted by a lot of single-celled organisms that are separately very simple, but together extremely complex. In some sense also the swarms of insects could be understood like a single entity.
Let us suppose now that every single part of similar entity is electronic, and, for some reason, that the experiment is beyond the control of the toy factory that developed it: the consequence is a catastrofic novel, The Reproductive System (a.k.a. Mechasm)⁽¹⁾, written with a lot of humor by John Sladek, in which the writer describes some mechanism that are just a bit more advanced version of the Molecule, a self-assemble modular robot developed by Keith Kotay and Daniela Rus at MIT:

Multiscale gigapixel photography

posted by @ulaulaman #photography #computer #science #technology

Sometimes, using the Android application Wondershare Panorama, I try to shot some panoramical photos. The result is good, but it could be better, like everything: indeed the actual camera (also in our smartphones) work near the fundamental limit of 1–10 megapixels for millimetre-scale apertures⁽¹⁾. In theory⁽²⁾ we are able to upgrade this limit. If we define $SW$ like the upper limit for the number of data channels which can be handled in parallel⁽²⁾, after some calculations, Adolf W. Lohmann found that without aberration, it would increase quadratically with the scaling factor $M$⁽²⁾. This result is different from the purely geometry result that states $SW$ indipendent from scaling.

That result cannot be realistic, otherwise, very long focal length lens systems would be fairly useless. In practice, the apertures of these long systems are reduced, more or less according to the empirical rule⁽²⁾

Now a team of researchers develop a new photographical device that can resolve at 50 gigapixels: AWARE-2 camera^{(1, *)}.
And below there are some examples of its capabilities⁽¹⁾: